Posts with heavy use of emoji and "leetspeak", while annoying, are not currently on the roadmap to be labeled by this service, so long as they can still technically be understood when read aloud. (Another party may consider writing a bot that could "un-Emojify" and "de-Leet" posts of those types.)